Series of periodic measurements or observations which are performed only in the intervention group. Appropriate for effect evaluation Maturation and testing are controlled in this design Failure to control for history (Only possible to be reduced when creating experimental isolation and constant conditions)

Home / Summaries / Health promotion disease prevention / intervention-control-evaluation

Evaluation - Research designs

Q: Randomized Controlled Trial (RCT)

High internal validity Good for efficacy studies (designed in order to make valid inferences about the intervention and its effect) Low external validity Less suitable for effectiveness studies (Aimed at evaluating the effect of interventions in real-life settings. These studies require designs in which the internal and external validity are more equally balanced).

Q: Non-equivalent control group design

Absence of randomisation Reduced internal validity (this design will never lead to the conclusion that the results are caused by the intervention and not by other factors) Equivalence between groups can be improved through matching or adjustment for differences in characteristics in the intervention and control group during statistical analyses. (Maturation, testing, history and regression to the mean can be controlled in this way for the most part) Well worth using when randomisation is impossible or undesirable

5 important questions on Evaluation - Research designs

Randomized Controlled Trial (RCT)

High internal validity

Good for efficacy studies (designed in order to make valid inferences about the intervention and its effect)

Low external validity

Less suitable for effectiveness studies (Aimed at evaluating the effect of interventions in real-life settings. These studies require designs in which the internal and external validity are more equally balanced).

CIT (community intervention trial)

Lower internal validity due to group randomization
Higher external validity because the intervention under evaluation better resembles the real-life situation

Non-equivalent control group design

Absence of randomisation

Reduced internal validity (this design will never lead to the conclusion that the results are caused by the intervention and not by other factors)
Equivalence between groups can be improved through matching or adjustment for differences in characteristics in the intervention and control group during statistical analyses. (Maturation, testing, history and regression to the mean can be controlled in this way for the most part)
Well worth using when randomisation is impossible or undesirable

Time series design

Series of periodic measurements or observations which are performed only in the intervention group.

Appropriate for effect evaluation
Maturation and testing are controlled in this design
Failure to control for history (Only possible to be reduced when creating experimental isolation and constant conditions)

One group pre-test post-test design

Due to the lack of a control group, this design is vulnerable for the effects of history, maturation, testing and regression to the mean
Appropriate for proces evaluation

The question on the page originate from the summary of the following study material:

Health promotion disease prevention

View summary